Maternal Mortality in a Nigerian Maternity Hospital

African Journal of Biomedical Research, Vol. 11 (2008); 267 – 273 ISSN 1119 – 5096 © Ibadan Biomedical Communications Group

Full Length Research Article Maternal Mortality in a Nigerian Maternity Hospital

*Olopade, F.E. and Lawoyin, T.O. Department of Community Health, College of Medicine,

Full-text available at University of Ibadan, Nigeria. http://www.ajbrui.com http://www.bioline.br/md http://www.ajol.com Despite recent focus on maternal mortality in Nigeria, its rates remain unacceptably high in Nigeria. A retrospective case-control study was carried out at Adeoyo Maternity Hospital, Ibadan between January 2003 and December 2004. This was to determine the maternal mortality ratio in a secondary health facility, to identify the causes of death and to assess factors associated with these deaths. The case files of all maternal deaths that occurred in the hospital during the two years period were retrieved and data extracted into a study proforma. Each maternal death was matched with three controls that delivered the same day and live around the same area of Ibadan. Bivariate analysis of the data was done with the cases and controls compared in relation to various risk factors. There were 8,724 lives births and 84 maternal deaths giving a MMR of 963/100,000 live births. The main causes of the death were haemorrhage 20 (23.8%) sepsis 20 (23.8%) and eclampsia 14(16.7%). Nine women (10.7%) died while pregnant, 11 (13.1%) died in labour, 52 (61.7%) died after delivery while 12 (14.3%) died from post-abortal complications. Most of the deaths due to post-partum heamorrhage Received: (66.7%) were seen in mothers over the age of 29 years while 64.2% of deaths due to February 2008 eclampsia were in women under the age of 25 years due to eclampsia occurred in nulliparous women and PPH was responsible for deaths in more women as their parity Accepted (Revised): increased from two. From bivariate analysis, factors significantly associated with June 2008 maternal death included unbooked status {Odd’s ratio OR=12.89 (95%C.I. _ 6.9 – 24.1) p<0.05}, grandmultiparity (parity>5) {OR=4.17 (95% C.I. – 1.53-11.45) p< 0.05} Published and prolonged labour (>18 hours) {OR=2.86, (95% C.I. 1.5 – 5.9) p< 0.05}. It is Septemebr 2008 necessary to improve the quality of emergency obstetric service as well as reduce the cost and to educate women of reproductive age in Ibadan on the importance of booking for antenatal care and family planning.

(Afr. J. Biomed. Res. 11: 267 - 273)

Key words: Maternal mortality, case-control study, maternal death, antenatal care, emergency obstetric care

*Corresponding author’s address: [email protected]

Abstracted by: African Index Medicus (WHO), CAB Abstracts, Index Copernicus, Global Health Abstracts, Asian Science Index, Index Veterinarius, Bioline International , African Journals online

African Journal of Biomedical Research 2008 (Vol. 11) / Olopade and Lawoyin

INTRODUCTION principal medical officers. There is no blood bank in the hospital and patients are required to procure Pregnancy and childbirth is a normal physiological blood from the State Transfusion Service or private process bringing a joyful experience to individuals laboratories around the hospital. The 16 bedded and families. However, in many parts of the world, labour ward records a delivery rate of 4,000-45,000 pregnancy and childbirth is a perilous journey, a annually. risky and potentially fatal experience for millions of All the maternal deaths that took place at the women especially in developing countries. This is in hospital between January 1, 2003 and December 31, spite of evidence which shows that motherhood can 2004 (a 2-year period) were identified. A maternal be safer for all women (De Browere, 1998), by a set death was defined as the death of a woman while of life-saving strategies that can work even in low pregnant or within six weeks of the end of resource settings (Shiffman, 2000). pregnancy irrespective of the duration or site of the The average maternal mortality rates in pregnancy, from any cause related to or aggravated developed countries is between 10-15/100,000 live by pregnancy and its management (WHO, 1992). births while developing countries record rates 100- Each maternal death was matched with three 200 times this number (Rosenfied, 1989). The controls that delivered on the same day and live problem of maternal deaths is worst in sub-Saharan within the same ward and similar socio-economic Africa with the maternal mortality rates there being area of Ibadan. The case files of all these women higher than anywhere else in the world (WHO, were retrieved from the Medical Records 2004). The situation in Nigeria is especially grave as Department and data extracted into a study we still record maternal mortality rates in the order proforma focusing on socio-demographic and of 800-1,000 per 100,000 live births (N.P.C. 2003) obstetric characteristics including age, parity, and thus rank among the nations with the highest occupational status, booking status, season of number of maternal deaths (WHO, 2004). Nigeria delivery and length of hospital stay before death as makes up only 1.7% of the total world population, well as cause of death where applicable. Data were but accounts for about 10% of the global estimate entered into a computer database using SPSS for maternal mortality (Grant, 1990). version 12 and statistical analysis was performed. A lot of work has been done on maternal Bivariate analysis was done using Epi-info 6. The mortality in tertiary health institutions, which have a odd’s ratios were generated to establish the high selection of complicated cases. In order to exposure as a risk factor for death. Maternal obtain a true picture of the epidemiology of mortality ratio was calculated as the number of maternal mortality in Ibadan, this study was carried maternal deaths per 100,000 live births (WHO, out in a secondary health facility to which primary 1992). health care facilities refer patients. Ethical approval was obtained from the Oyo State Ethical Review Committee and the MATERIALS AND METHODS management of the facility used.

This is a retrospective case - control study of RESULTS maternal deaths which was conducted in Adeoyo Maternity Hospital. It is an 80-year-old state-owned During the study period, a total of 84 maternal General Hospital in Ibadan city, Nigeria. It is highly deaths and 8,724 live births were recorded at the patronized by Ibadan residents especially those of hospital. Table 1 shows a review of the socio- low and middle socio-economic status. It also serves demographic characteristics of the cases and as a referral center for many primary health centers controls and reveals no significant differences and private clinics within Ibadan and its environs. between them except for parity and booking status. The Obstetrics and Gynaecology department of Most of the women (70.2% of the cases and 72% of the hospital has 2 consultants and about 10 doctors the controls) were unskilled, mainly petty traders. including medical officers, senior medical and The study maternal mortality ratio was

268 Maternal mortality in Nigeria African Journal of Biomedical Research 2008 (Vol. 11) / Olopade and Lawoyin calculated as 963/100,000 live births. Direct increased significantly as the parity of the women obstetric deaths were 75 (89.3%) while indirect increased; no deaths in women with parity below obstetric deaths were 9 (10.7%). About 61.9% of the 2, 23.5% of deaths in women with parity of 2, and maternal deaths occurred post-partum, 13.1% intra- 36.4% of women with parity ! 5. partum and 10.7% ante-partum while 14.3% of the women died from post-abortal complications. The main causes of death were haemorrhage 20 (23.8%), sepsis 20 (23.8%), eclampsia 14 (16.7%), post- abortal complications 11 (13.1%) and obstructed labour 10 (11.9%) (Figure 1).

Table 1: Socio-Demographic Pattern of Cases Characteristics Cases N (%) AGE (YEARS) 15-19 6 (7.1) 20-24 15 (17.9) 25-29 35 (41.7) 30-34 20 (23.8) ! 35 8 (9.5) PARITY 0 21 (25) 1 12 (14.3) Figure 1: Causes of Maternal Death at Adeoyo Maternity 2 17 (20.2) Hospital (2003-2004) 3 11 (13.1) *Includes ante-partum haemorrhage (APH) and post-partum 4 10 (11.9) haemorrhage (PPH) **Anaesthetic deaths, Pulmonary TB. ! 5 11 (13.9) Not recorded 2 (2.4) Length of admission and Cause of Death: Most OCC. STATUS of the eclamptic deaths (85%), deaths due to Higher (Prof. and Skilled 6 (7.2) obstructed labour (80%) and post-partum Lower (Unskilled) 59 (70.2) haemorrhage (73.3%) occurred within 24 hours of Unemployed 19 (22.6) admission, but at least 40% of deaths due to sepsis BOOKING STATUS occurred after 72 hours of admission. Booked 23 (27.4) Booking Status and Cause of Death: Deaths due Unbooked 61 (72.6) to sepsis (75%), eclampsia (71.4%), ante-partum SEASON OF DELIVERY haemorrhage (100%) and obstructed labour (80%) Dry (November—March) 36 (42.9) occurred more in unbooked women while almost Wet (April—October) 48 (57.1) equal proportions of booked and unbooked women died due to post-partum haemorrhage and Table 2 compares the obstetric history of the cases anaemia. with their various causes of death. Season of Delivery and Cause of Death: Significantly more eclamptic deaths were seen in Age and Cause of Death: Most of the deaths due the rainy season (x2 = 5.14 1 d.f. p = 0.023) as well to post-partum haemorrhage (66.7%) occurred in as deaths due to obstructed labour (x2 = 3.2 1 d.f. women over 29 years of age while 64.2% of the p = 0.074) and post abortal complications (x2 = 6.6 deaths due to eclampia were in women under 25 1 d.f. p = 0.01). All deaths due to anaemia years of age. occurred during the rainy season. From bivariate Parity and Cause of Death: Of the deaths due to analysis, the risk factors identified to be associated eclampia, 71.4% occurred in nulliparous women with maternal death (See Table 3) were while post-partum haemorrhge as a cause of death grandmultiparity (parity !5), prolonged labour (!

Maternal mortality in Nigeria 269 African Journal of Biomedical Research 2008 (Vol. 11) / Olopade and Lawoyin

18hours), raised blood pressure on admission (≥ figure for Nigeria, which is 800/100,000 live 160/90mmHg), lower occupational status and lack births. However, it is lower than what is obtained of antenatal care. The age of the mothers was not a in studies done in teaching hospitals around the significant risk factor for maternal death in this nation which put the figure between 1,000 and study. Teenage mothers (age <20yrs) and older 3,0000/100,000 live births (Oladapo et al., 2006; mothers (≥ 35yrs) were not at a statistically Daramola et al., 2004). This is believed to be a significant risk of death. truer picture of the MMR in Ibadan since the study was carried out in a secondary health facility DISCUSSION which does not have such a high selection of complicated cases as in the teaching hospitals and The study maternal mortality ratio (MMR) of receives referred cases from the grassroots through 963/100,000 live births is higher than the national the primary health care system.

Table 2: Obstetric History and Causes of Death Obst. Post abortal Characteristics PPH APH Eclampsia Sepsis Anaemia Others Labour Comp. AGE (Yrs) 15-19 0 0 4 1 0 1 0 0 20-24 1 0 5 3 1 2 1 2 25-29 4 5 1 12 4 4 4 1 30-34 8 0 2 3 4 2 0 1 ≥ 35 2 0 2 1 1 2 0 0 TOTAL 15 5 14 20 10 11 5 4 PARITY Nulliparous 0 1 10 5 1 1 2 1 1 0 0 0 4 4 2 1 1 3 4 1 3 1 1 3 2 2 3 & 4 7 2 1 7 2 4 0 0 5 4 1 0 3 2 1 0 0 TOTAL 15 5 14 20 10 11 5 4 LENGTH OF ADMISSION (Hrs) 0-24 11 3 12 7 8 7 0 2 25-48 2 0 1 2 2 1 2 0 49-72 1 1 0 3 0 2 1 0 72 1 1 1 8 0 1 2 2 TOTAL 15 5 14 20 10 11 5 4

BOOKING STATUS Booked 7 0 4 5 2 0 3 2 Unbooked 8 5 10 15 8 11 2 2 TOTAL 15 5 14 20 10 11 5 4 SEASON OF DELIVERY Dry (Nov.-Mar.) 9 2 4 9 3 3 0 1 Wet (Apr.-Oct.) 6 3 10 11 7 8 5 3 TOTAL 15 5 14 20 10 11 5 4

270 Maternal mortality in Nigeria African Journal of Biomedical Research 2008 (Vol. 11) / Olopade and Lawoyin

Table 3: Risk factors Associated with Maternal Death Risk Factor* Odd’s Ratio 95% C.I. (Min.) 95% C.I. (Max.) p- value Age < 20Yrs (20-34 Yrs) 1.06 0.36 3.02 0.9 Age > 34 Yrs (20-34 Yrs) 1.79 0.66 4.77 0.31 Nulliparity (Parity 1-4) 0.75 0.04 1.37 0.31 Grandmultiparity (Parity 1-4) 4.17 1.53 11.45 0.0026 Long labour (<18 Hours) 2.86 1.39 5.9 0.0016 Raised Systolic B.P. (<160mmHg) 5.44 2.15 13.96 0.000023 Raised Diastolic B.P. (<90 mmHg) 2.64 1.41 4.93 0.00079 Dry season (Wet season) 1.33 0.77 2.27 0.269 Higher occupational status (Lower Occ. Status) 0.34 0.12 0.87 0.01 No antenatal care (At least one ANC visit) 12.89 6.94 24.1 0.000001 3-5 yrs birth interval (> 6yrs interval) 0.3 0.08 1.14 0.073 * Reference category in parenthesis

Obstetric haemorrhage and sepsis together who are unable to afford these are usually those accounted for almost half (47.6%) of all maternal that die. deaths in the study period. This reflects an Eclampsia is reported to be the most difficult inadequacy of the emergency obstetric response of the obstetric emergencies to prevent and which is common to other developing countries manage as well as the least understood (Howson (Matthews, 2005). Although haemorrhage may be et. al., 1995). Although it is not a common cause an important cause of death in some developed of death in industrialized countries, it is very countries, most of the deaths are due to ruptured common in developing countries (Duley 1992; ectopic pregnancy (Jacobs et. al., 1998) and not Moodley and Daya, 1993). In this study, eclampsia post partum haemorrhage (PPH) as we have here. occurred mostly in primigravidae (71.4%) and PPH was found to be responsible for more deaths young mothers (< 25 years) (64.3%) and over 70% in older (≥ 30 years), grandmultiparous (parity ≥ of the deaths occurred in unbooked women. This 5) women, which may be due to an increased is similar to findings by Onuh and Aisien (2004) incidence of uterine atony with increasing age and who reported that eclampsia is most frequently parity. About 70% of these deaths occurred within seen in the first pregnancy and closely linked to 24 hours of admission thus revealing that nothing lack of antenatal care. Antenatal care is short of rapid medical response including indispensable to maternal health as it is needed for immediate arrest of haemorrhage and blood early detection and management of hypertension transfusion would save a woman with this in pregnancy. Sadly, it is reported that only 63% condition. The absence of a blood bank within the of women in Africa receive antenatal care and hospital premises may have contributed to some of fewer still deliver in a health care facility (WHO, these deaths. 1997). This study also reports that 85.7% of the Sepsis, as a cause of death was not modified eclamptic deaths occurred within 24 hours of by age or parity of the mothers, though further hospital admission, thus suggesting either late statistical analysis needs to be done to ascertain reporting by the patients or inadequate supportive this. However, it was noted that most of these therapy at the hospital. deaths occurred after 48 hours of admission, thus Deaths due to obstructed labour occurred indicating a longer course and a window of mainly in women aged between 25-34 years (80%) opportunity for intervention which was hardly and in women that have delivered at least a child utilized. Puerperal sepsis which is commonly before. This may be due to delay on the part of the complicated by anaemia can be adequately patients and their relatives as they may not readily managed by aggressive antibiotic administration consent to operative delivery since she had and blood transfusion when necessary. Patient delivered vaginally in the past. Also, obstructed

Maternal mortality in Nigeria 271 African Journal of Biomedical Research 2008 (Vol. 11) / Olopade and Lawoyin

labour in multigravidae is more likely to lead to a In general, women who had a lower occupational ruptured uterus which has a high case fatality rate status were more likely to die than those with a (Rush, 2000; Ahmed et al., 2004). higher status. This could be because the poorer Over 90% of the maternal deaths due to post ones are not able to afford the necessary abortal complications were in women aged over management for their particular condition. 20yrs and those who had at least one child. This is In conclusion, emergency obstetric care needs interesting to note and is in agreement with the to be provided free or at a WHO (1998) conclusion that most women seeking highly subsidized rate for all pregnant women abortion are married or live in stable unions and using this facility. It is also essential that a referral already have children; they seek abortion primarily center such as the one in which the study was to limit the size of their family or to space births. carried out should have a blood bank within its Most deaths from post abortal sepsis in an Indian premises. Widespread education of women in the study were also found to be in married women reproductive age in Ibadan, as well as their (Chhabra et. al., 2005). This is alarming because it spouses on the importance of regular antenatal points out the extreme importance of family care and family planning is essential in order to planning which is not fully accepted in this reduce maternal deaths in this environment. environment. It is reported that less than 10% of Nigerian women in the reproductive age group are REFERENCES on modern contraception. There is the common belief that only promiscuous women use Ahmed Y., Shehu C.E., Nwobodo E.I. and Ekole B.A. contraception, so most husbands do not allow their (2004). Reducing maternal mortality from ruptured uterus wives to use them. A massive campaign on family – the Sokoto intiative. African Journal of Medical planning targeting the men and trying to change Science 33(2): 135-138. Begum, S., Nisa, A. and Begum, I. (2003). Analysis of this mentality is needed to stop this ugly trend and maternal mortality in a tertiary care hospital to determine save the lives of women. causes and preventable factors. J. Ayub.Med. Coll. Grandmultiparous women (parity ≥5) had at Abboattabad. 15:2. least a four-fold increased risk of death when Chhabra S., Kaipa A. and Kanani A. (2005). compared with those with parity <5. This is Reduction in Maternal mortality due to sepsis. Journal of supported by various studies around the world Obstetrics and Gyaecology 25(2): 140-142. (Garenne et al., 1997; Begum et al., 2003). These Daramola, A.O., Banjo, A.A. and Elesha, S.O. (2004). women need not have died if they had limited their Maternal deaths in the Lagos University Teaching family to less than five. Mothers who were Hospital: a ten-year review (1989 – 98). Nigerian unbooked for antenatal care also had a higher risk Postgraduate Medical Journal. 11(4): 274 – 278. De Browere V. (1998). Strategies for reducing maternal of death. This may be because they only present at mortality in developing countries: What can we learn the hospital when they have severe, sometimes life from the history of the industrialized West? Tropical threatening complications (Harrison, 1997). Medicine and International Health 3(10): 771-782. Similarly, women who had prolonged labour (≥ Duley L. (1992). Maternal mortality associated with 18hours) faced a 3-fold increased risk of maternal hypertensive disorders of pregnancy in Africa, Asia, death. Indeed they are more exposed to infection, Latin America and the Caribbeans. British Journal of maternal exhaustion and chances of instrumental Obstetric and Gynaecology 99:547-553. delivery all of which increase their risk of death. Garenne, M., Mbaye, K., Bah, M.O. and Correa, P. A raised blood pressure on admission (≥ (1997). Risk factors for maternal mortality: a case-control 160/90mmHg) is usually an ominous sign as such study in dakar hospitals (Senegal). African Journal of Reproductive Health 1(1): 14-24. women were 2-5 times more likely to die than Grant J.P. (1990). State of the World’s Children, New those with normal blood pressure. It is important York Oxford University Press, 1990. that such women are adequately cared for in order Harrison, K.A. (1997). Maternal mortality in Nigeria: to prevent them from having convulsions and the real issues. African Journal of Reproductive Health 1 putting their lives and that of their babies at risk. (1): 7-13. Howson C.P., Harrison P.F., Hana D. and Law M.

272 Maternal mortality in Nigeria African Journal of Biomedical Research 2008 (Vol. 11) / Olopade and Lawoyin

(eds). (1996). In her lifetime – female morbidity and 262(3): 376-379. mortality in sub-Saharan Africa. National Academy Rush D. (2000). Nutrition and Maternal Mortality in the Press, Washington DC, USA, 1996. Developing World. American Jorunal of Clinical Jacob S., Bloebaum L., Shah G. and Varner M.W. Nutrition 72(1): 22-240. (1998). Maternal Mortality in Utah, Obstetrics and Shiffman J. (2000). Can poor countries surmount high Gynaecology 91: 187-191. maternal mortality? Studies in Family Planning, 31(4): Matthews M. (2005). Improving Maternal and Child 274-289. Survival in India. Indian Journal of Medical Research World Health Organization (1992). International 121: 621-627. Classification of Diseases and health-Related Problems in Moodley J. and Moodley D. (2005). Management of the 10th Revision, Volume 2. World Health Organization, Human Immuno-Deficiency Virus Infection in Geneva. Pregnancy. Best Pract. Clin. Obstet. Gynaecol, 19(2): World Health Organization (1997). Coverage of 169-183. Maternal Care: A Listing of Available Information, 4th National Population Commission (2003). Nigeria edition, WHO, Geneva.edition, WHO, Geneva. Demographic and Health Survey (2003). IRD Macro World Health Organization (1998). Abortion: A International USA. Tabulation of Available Information, 3rd edition. World Oladapo, O.T., Lamina, M.A., and Fakoya, T.A. Health Organization, Geneva. (2006). Maternal deaths in Sagamu in the new World Health Organization (1999). Reduction of millennium: a facility-based retrospective analysis. Maternal Mortality: A Joint Bio.Med Central Pregnancy and Childbirth 6:6 WHO/UNFPA/UNICEF/World Bank Statement, WHO, Onuh S.O. and Aisien A.O. (2004). Maternal and Fetal Geneva. Outcome in Eclamptic Patients in Benin City, Nigeria. J. World Health Organization (2004). Maternal Mortality Obstet. Gynaecology 24(7): 765-768. in 2000: Estimates Developed by WHO, UNICEF and Rosenfield A. (1989). Maternal Mortality in Developing UNFPA. World Health Organization, Geneva. Countries. An ongoing but neglected tragedy. JAMA

Maternal mortality in Nigeria 273